From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail144.messagelabs.com (mail144.messagelabs.com [216.82.254.51]) by kanga.kvack.org (Postfix) with ESMTP id 82E226B00AA for ; Wed, 1 Dec 2010 15:16:38 -0500 (EST) Received: from mail-iw0-f169.google.com (mail-iw0-f169.google.com [209.85.214.169]) (authenticated bits=0) by smtp1.linux-foundation.org (8.14.2/8.13.5/Debian-3ubuntu1.1) with ESMTP id oB1KG9UU023912 (version=TLSv1/SSLv3 cipher=RC4-MD5 bits=128 verify=FAIL) for ; Wed, 1 Dec 2010 12:16:09 -0800 Received: by iwn41 with SMTP id 41so388505iwn.14 for ; Wed, 01 Dec 2010 12:16:08 -0800 (PST) MIME-Version: 1.0 In-Reply-To: <1330724443.975931291231775834.JavaMail.root@zmail06.collab.prod.int.phx2.redhat.com> References: <919384632.877731291171499343.JavaMail.root@zmail06.collab.prod.int.phx2.redhat.com> <1330724443.975931291231775834.JavaMail.root@zmail06.collab.prod.int.phx2.redhat.com> From: Linus Torvalds Date: Wed, 1 Dec 2010 12:15:48 -0800 Message-ID: Subject: Re: oom is broken in mmotm 2010-11-09-15-31 tree? Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Sender: owner-linux-mm@kvack.org To: CAI Qian Cc: linux-mm , Michel Lespinasse , Rik van Riel , Wu Fengguang , "H. Peter Anvin" List-ID: On Wed, Dec 1, 2010 at 11:29 AM, CAI Qian wrote: >> >> Hi, just a head-up. When testing oom for this tree, my workstation is >> immediately having no response to ssh, Desktop actions and so on apart >> from ping. I am trying to bisect but looks like git public server is >> having problem. > > This turned out that it was introduced by, > > =A0d065bd810b6deb67d4897a14bfe21f8eb526ba99 > =A0mm: retry page fault when blocking on disk transfer > > It was reproduced by: > 1) ssh to the test box. > 2) try to trigger oom a few times using a malloc program there. Interesting. That commit is not supposed to make any semantic difference at all. And even if we do end up in the retry path, the arch/x86/mm/fault.c code is very explicitly designed so that it retries only _once_. Michel, any ideas? I could see problems with the mmap_sem if VM_FAULT_OOM is set at the same time as VM_FAULT_RETRY, but I can't see how that could ever happen. Anybody? CAI, can you get any output from sysrq-W when this happens? Linus -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Fight unfair telecom policy in Canada: sign http://dissolvethecrtc.ca/ Don't email: email@kvack.org